On 2018 Feb 02, Christopher Korch commented:

In this article by Binder et al. in PLOS ONE (Binder NK, 2016), the authors describe using the cell line ECC-1 as a model of endometrial tissue to study embryo implantation. Using recombinant human placental growth factor they found that treatment of ECC-1 cells increased their cellular adhesion to fibronectin-coated tissue culture plates. I wish to point out two major concerns about the authenticity of this cell line.

First, two of the three references which they cite for having characterized ECC-1 are not correct. References 23 and 24 refer to the cell line HES, which my colleagues and I showed in 2012, was actually the HeLa subline WISH; i.e., a cervical adenocarcinoma cell line (Korch et al. Korch C, 2012) and our finding was confirmed in 2014 by the originator of HES (Kniss & Summerfield Kniss DA, 2014). WISH was shown by SM Gartler in 1966-1970 to be HeLa cells (Gartler SM, 1967, Gartler SM, 1968, Auersperg N, 1970) and this has been confirmed numerous times thereafter (e.g., by Nelson-Rees during the 1970s-1980s Nelson-Rees WA, 1980, Nelson-Rees WA, 1981; Lavappa at the ATCC in 1978 Lavappa KS, 1978; and Masters et al. 2001, Sandler AN, 1992). No authentic sample of WISH is known to exist (see cell register of misidentified cell lines at the websites of the International Cell Line Authentication Committee (http://ICLAC.org) and of the Expasy Cellosaurus (https://web.expasy.org/cellosaurus/). In fact, ECC-1 is also listed on both websites as a misidentified cell line.

Next, the authors do not indicate whether they genetically authenticated their sample of ECC-1 (currently the method of reference is by STR genotyping). This cell line was developed by PG Satyaswaroop's group at the Hershey Medical Center in about 1985-1987 when it was established from a xenograft sample of the endometrial tumor EnCa101, which was being maintained by passaging in mice. The cell line ECC-1 was deposited at the ATCC by Bruce Lessey, who had received it from Dr. Sayaswaroop. It was STR genotyped by the ATCC. As described earlier by me and my colleagues (Korch C, 2012) from STR genotyping of numerous samples of ECC-1 and Ishikawa cells from various sources and comparison to the STR profile of the ATCC sample of this cell line, ECC-1 was found to be one of three cell lines - a derivative of the endometrial cell line Ishikawa developed by Nishida (see Nishida M, 2002 for history and dissemination of various subclones), the breast cancer cell line MCF-7, or a mixture of Ishikawa and MCF-7 cells. A complicating fact is that ISHIKAWA / ECC-1 is an MSI unstable cell line, giving rise to variable STR genotypes.

In an attempt to determine the expected STR profile of ECC-1 / EnCa101, the Hershey Medical Center was approached, but no samples of the original tumor (paraffin block, etc) could be found as the Satyaswaroop lab had been closed in 2002. Furthermore, none of the ECC-1 or Ishikawa samples matched any of the four samples of xenografts of the tumor EnCa101. This tumor has been maintained by VC Jordan since about 1987 (Gottardis MM, 1988). The earliest tumor sample that I could obtain of EnCa101was from 1987 and was found to be genetically unrelated to the three tumor xenograft samples of EnCa101 from 2009 (unpublished data) and the STR genotypes of these four samples did match the profile of any known cell line in several databases.

Therefore, these results using the cell line ECC-1 should be re-interpreted and used cautiously, because (1) this cell line is a misidentified cell line as shown earlier (Korch C, 2012) and is listed as a misidentified cell line on the ICLAC and Cellosaurus websites; (2) the true identity of the sample of ECC-1 that was used by the authors was not determined, (3) its provenance is confusing because two of the three references for its characterization are incorrect and refer to a different cell line, and (4) since the genetic identity of ECC-1 cell line was not checked, it could be one of at least five different cultures (three variants of Ishikawa and breast cancer cell line and a mixture of Ishikawa and MCF-7).

Hopefully, this will encourage others to authenticate their cell lines. Below are some suggestions of how this problem could be avoided in the future. They are based on ideas put forth earlier by Dr. Amanda Capes-Davis in a PubMed Commons Comment on another article of concern (see Xu HT, 2016).

Authors & Reviewers could use the aforementioned resources since: • STR genotyping is effective for authentication of human cell lines and is the consensus method for comparison of human cell line samples (American Type Culture Collection Standards Development Organization Workgroup ASN-0002., 2010, American Type Culture Collection Standards Development Organization Workgroup ASN-0002., 2010); and

• Checklists for the identity of cell lines used in manuscripts and grant applications are readily available at http://iclac.org/resources/cell-line-checklist/ and through Expasy Cellosaurus https://web.expasy.org/cellosaurus/.

Journals, their Editors, and Funding Organizations could: • Implement more stringent criteria for publication and funding of research because encouragement of authentication testing, although a step forward, is insufficient to stop use of misidentified cell lines.

• Develop an authentication policy for sake of reproducible research. For example, to meet this need the NIH has recently implemented requirements for authentication of key resources as part of grant applications (e.g. see NOT-OD-16-011, NOT-OD-16-012).

• Require testing using an accepted method of cell line authentication, which is effective as illustrated by the cell line authentication policy of the International Journal of Cancer (Fusenig NE, 2017). If the authors, the reviewers, or the journal editors had followed PLOS ONE's own guidelines (http://journals.plos.org/plosone/s/submission-guidelines#loc-cell-lines) and checked for the identity of this cell line on either the ICLAC website (http://iclac.org/databases/cross-contaminations/) or on the Cellosaurus website (https://web.expasy.org/cellosaurus/), the journal would have detected and avoided this problem prior to publication.

• Regularly review the efficacy of their policy on cell authentication testing to see whether it is adequate (as was done by the International Journal of Cancer), especially in light of such examples having been published by a journal that requires authentication of cell lines prior to submission of manuscripts.

On 2018 Feb 02, Christopher Korch commented:

In this article by Binder et al. in PLOS ONE (Binder NK, 2016), the authors describe using the cell line ECC-1 as a model of endometrial tissue to study embryo implantation. Using recombinant human placental growth factor they found that treatment of ECC-1 cells increased their cellular adhesion to fibronectin-coated tissue culture plates. I wish to point out two major concerns about the authenticity of this cell line.

First, two of the three references which they cite for having characterized ECC-1 are not correct. References 23 and 24 refer to the cell line HES, which my colleagues and I showed in 2012, was actually the HeLa subline WISH; i.e., a cervical adenocarcinoma cell line (Korch et al. Korch C, 2012) and our finding was confirmed in 2014 by the originator of HES (Kniss & Summerfield Kniss DA, 2014). WISH was shown by SM Gartler in 1966-1970 to be HeLa cells (Gartler SM, 1967, Gartler SM, 1968, Auersperg N, 1970) and this has been confirmed numerous times thereafter (e.g., by Nelson-Rees during the 1970s-1980s Nelson-Rees WA, 1980, Nelson-Rees WA, 1981; Lavappa at the ATCC in 1978 Lavappa KS, 1978; and Masters et al. 2001, Sandler AN, 1992). No authentic sample of WISH is known to exist (see cell register of misidentified cell lines at the websites of the International Cell Line Authentication Committee (http://ICLAC.org) and of the Expasy Cellosaurus (https://web.expasy.org/cellosaurus/). In fact, ECC-1 is also listed on both websites as a misidentified cell line.

Next, the authors do not indicate whether they genetically authenticated their sample of ECC-1 (currently the method of reference is by STR genotyping). This cell line was developed by PG Satyaswaroop's group at the Hershey Medical Center in about 1985-1987 when it was established from a xenograft sample of the endometrial tumor EnCa101, which was being maintained by passaging in mice. The cell line ECC-1 was deposited at the ATCC by Bruce Lessey, who had received it from Dr. Sayaswaroop. It was STR genotyped by the ATCC. As described earlier by me and my colleagues (Korch C, 2012) from STR genotyping of numerous samples of ECC-1 and Ishikawa cells from various sources and comparison to the STR profile of the ATCC sample of this cell line, ECC-1 was found to be one of three cell lines - a derivative of the endometrial cell line Ishikawa developed by Nishida (see Nishida M, 2002 for history and dissemination of various subclones), the breast cancer cell line MCF-7, or a mixture of Ishikawa and MCF-7 cells. A complicating fact is that ISHIKAWA / ECC-1 is an MSI unstable cell line, giving rise to variable STR genotypes.

In an attempt to determine the expected STR profile of ECC-1 / EnCa101, the Hershey Medical Center was approached, but no samples of the original tumor (paraffin block, etc) could be found as the Satyaswaroop lab had been closed in 2002. Furthermore, none of the ECC-1 or Ishikawa samples matched any of the four samples of xenografts of the tumor EnCa101. This tumor has been maintained by VC Jordan since about 1987 (Gottardis MM, 1988). The earliest tumor sample that I could obtain of EnCa101was from 1987 and was found to be genetically unrelated to the three tumor xenograft samples of EnCa101 from 2009 (unpublished data) and the STR genotypes of these four samples did match the profile of any known cell line in several databases.

Therefore, these results using the cell line ECC-1 should be re-interpreted and used cautiously, because (1) this cell line is a misidentified cell line as shown earlier (Korch C, 2012) and is listed as a misidentified cell line on the ICLAC and Cellosaurus websites; (2) the true identity of the sample of ECC-1 that was used by the authors was not determined, (3) its provenance is confusing because two of the three references for its characterization are incorrect and refer to a different cell line, and (4) since the genetic identity of ECC-1 cell line was not checked, it could be one of at least five different cultures (three variants of Ishikawa and breast cancer cell line and a mixture of Ishikawa and MCF-7).

Hopefully, this will encourage others to authenticate their cell lines. Below are some suggestions of how this problem could be avoided in the future. They are based on ideas put forth earlier by Dr. Amanda Capes-Davis in a PubMed Commons Comment on another article of concern (see Xu HT, 2016).

Authors & Reviewers could use the aforementioned resources since: • STR genotyping is effective for authentication of human cell lines and is the consensus method for comparison of human cell line samples (American Type Culture Collection Standards Development Organization Workgroup ASN-0002., 2010, American Type Culture Collection Standards Development Organization Workgroup ASN-0002., 2010); and

• Checklists for the identity of cell lines used in manuscripts and grant applications are readily available at http://iclac.org/resources/cell-line-checklist/ and through Expasy Cellosaurus https://web.expasy.org/cellosaurus/.

Journals, their Editors, and Funding Organizations could: • Implement more stringent criteria for publication and funding of research because encouragement of authentication testing, although a step forward, is insufficient to stop use of misidentified cell lines.

• Develop an authentication policy for sake of reproducible research. For example, to meet this need the NIH has recently implemented requirements for authentication of key resources as part of grant applications (e.g. see NOT-OD-16-011, NOT-OD-16-012).

• Require testing using an accepted method of cell line authentication, which is effective as illustrated by the cell line authentication policy of the International Journal of Cancer (Fusenig NE, 2017). If the authors, the reviewers, or the journal editors had followed PLOS ONE's own guidelines (http://journals.plos.org/plosone/s/submission-guidelines#loc-cell-lines) and checked for the identity of this cell line on either the ICLAC website (http://iclac.org/databases/cross-contaminations/) or on the Cellosaurus website (https://web.expasy.org/cellosaurus/), the journal would have detected and avoided this problem prior to publication.

• Regularly review the efficacy of their policy on cell authentication testing to see whether it is adequate (as was done by the International Journal of Cancer), especially in light of such examples having been published by a journal that requires authentication of cell lines prior to submission of manuscripts.